Since the post-intensive care syndrome (PICS) was first described, multiple publications have addressed it, either in its entirety or partially. However, the purpose of this approach is not to comprehensively examine PICS, but to draw attention on the epidemiology of mental disorders. Although this aspect has been extensively addressed by the international scientific medical literature, uncertainties abound that can only be resolved through more rigorous analysis.

Three meta-analyses were published to assess the prevalence of symptoms of anxiety,1 depression,2 and post-traumatic stress disorder (PTSD)3 in survivors of critical illness from 2015 through 2016. The conclusions were that 32%–40%, 30%, and 22% of the patients, respectively, showed symptoms of anxiety,1 depression,2 or PTSD,3 at the 1-year follow-up. All of them exhibited an exaggerated heterogeneity among the results of the studies included. In most of these, data were obtained using surveys for symptom detection without in-person evaluations, a limitation that was noted in all 3 meta-analyses.

Both this and other methodological limitations (small sample sizes, lack of background knowledge, heterogeneity of the patients included, etc.) have persisted up to this day, and the results continue to be characterized by their variability. Although many emphasize the high prevalence of symptoms, the limitations affecting the results are also recognized. Among these, the exclusive assessment of symptoms using scales (telephone or mail-based in some cases) without specialized in-person evaluations is repeatedly acknowledged, an approach that lacks the necessary methodological robustness for the accurate identification of mental disorders (Table 1).4–13

In former studies whose objective was to determine their prevalence after the ICU admission by specialized personnel (although with different methodologies), more homogeneous results have been reported, ranging from 12%9 to 17%8 for psychiatric diagnoses. In a recent study13 conducted in 3 national centers, we evaluated a total of 186 COVID-19 survivors who had been mechanically ventilated and concluded that 31% presented psychiatric disorders 3 months after hospital discharge. Although the assessments were conducted in person, these data were obtained from the scales used in the Intensive Care Medicine clinic (which means that we were not unaware of the previously mentioned limitations). The follow-up of patients referred to the Psychiatry team (unpublished data) revealed that only 15% of them were diagnosed with a mental disorder.

Assessment scales are used to measure and quantify symptoms, behaviors, or mental traits. They are transformed into numerical values that indicate the severity of specific symptoms, but they do not constitute a definitive diagnosis. Although it has been argued that some of them are suitable detection tools due to their sensitivity and specificity, both can vary depending on various factors such as the thresholds used to define the presence of symptoms, study population, context, etc. and should not replace specialized in-person evaluation.

It is unquestionable that mental disorders often follow a critical illness, either as new diagnoses or exacerbations of pre-existing psychiatric conditions (an aspect that has not been rigorously addressed in the scientific medical literature to date). However, the extensive bibliography that has based its conclusions on the results of surveys for symptom detection without in-person validation can lead to psychiatric overdiagnosis, resulting in alarming speculation not only in the scientific community but also in society, often fueled by sensational headlines in the general press. Future epidemiological studies must be founded on the necessary methodological rigor to obtain reliable results and avoid overestimated data that can trigger confusion, promote misinterpretations, and feed speculation.